Temporal inabilities and decision-making capacity in depression

We report on an interview-based study of decision-making capacity in two classes of patients suffering from depression. Developing a method of second-person hermeneutic phenomenology, we articulate the distinctive combination of temporal agility and temporal inability characteristic of the experience of severely depressed patients. We argue that a cluster of decision-specific temporal abilities is a critical element of decision-making capacity, and we show that loss of these abilities is a risk factor distinguishing severely depressed patients from mildly/moderately depressed patients. We explore the legal and clinical consequences of this result.

the ways in which severe depression can undermine the ability of depressed individuals to deliberate and make decisions for themselves. Decision-making capacity (or simply 'mental capacity') 1 is an increasingly important concept in law and clinical practice; medical and legal professionals are now regularly called upon to offer professional judgements as to its presence or absence. Psychiatrists often comment on the difficulty of assessing decision-making capacity in people with depression; better understanding of the phenomenology of depression is imperative for refining clinical and legal techniques for such determinations. The results of the present phenomenological study of depression show that decision-making capacity is partly constituted by forms of temporal agility and can be compromised by temporal inability. We discuss theoretical and practical implications, especially in relation to contemporary legal models of decision-making capacity.
A link between depression and temporal experience has been widely reported and discussed in the psychiatric literature, while temporality has long been a central theme of phenomenological research; a handful of recent studies have undertaken to bring these two research traditions together. The present project draws on this rich body of existing research, while seeking to advance it in two ways. Our comparative empirical study provides data whereby existing phenomenological models of temporal experience in depression can be assessed, refined and (where necessary) corrected. And we draw out specific consequences of this psychiatric and phenomenological research to navigate the concrete medico-juridical challenge of assessing decision-making capacity. We argue that responsible assessment of decision-making capacity among depressed patients requires application of a phenomenologically informed understanding of the combination of temporal agility and temporal inability that is at work in depressive hopelessness.

Design of the study
The methodology of the study was a principled hybrid, integrating phenomenological and psychiatric elements with standard techniques from qualitative research protocols. The core strategy involved a technique of second-person hermeneutic phenomenology. The strategy was to use a series of clinical interviews with patients in order to gain insight into the texture of the patient's experience of making decisions under conditions of psychiatric disorder. The interviews were recorded and transcribed and subsequently 'coded' using line-by-line gerundive coding techniques (Charmaz 2006). The recordings, coded transcripts and contextual information were then submitted to close interpretative scrutiny. In no instance did we simply assume that the transcripts could be taken at face-value; see Section 4, below. Instead, we used the data from the interviews as evidence from which hypotheses regarding the patient experience could be generated; we then probed these hypotheses both in subsequent interviews and in further analysis of recordings and transcripts. We used techniques of grounded theory (Charmaz 2006) to assist in the construction of higher-order codes and in the thematisation of our data. Excerpts from the clinical interviews cited in this article are drawn from a large corpus of interview data.
Second-person phenomenology and standard qualitative research methodology were supplemented by canonical first-person (or 'reflective') phenomenology. We studied the classical investigations of temporal experience from the so-called 'phenomenological tradition', and we drew on our own experience of the temporal structures at work in our own decision-making. Third-person or 'objective' methodology was also employed in that we selected cases of depression according to a public psychiatric classification system (ICD-10), using traditional psychiatric assessment tools. Such measures relate to the body of medical knowledge on depression (causation, treatment, prognosis, etc.) and form admissible evidence in courts in legal determinations of 'true mental disorder'.
The interviews were organised around two decisions: We asked patients to report on the decision that they had made to participate in the research study itself; we then explored a pending decision they faced concerning treatment options for their depression. Seven patients with severe depression (D1-7) participated in the study. Because depression is understood as a graded phenomenon, and because doubts about decisionmaking capacity tend to arise only in cases where depression is severe, we also wanted a contrast sample where depression would be present but where lack of decision-making capacity was unlikely. We therefore sampled six cases of people (N1-6) suffering depression at the other end of the severity spectrum ('mild') to study in comparison to the severe cases. All patients were recruited from a pool of people in contact with the South London and Maudsley National Health Service Foundation Trust, a mental health service provider for people in South East London, UK. We include further technical details regarding the study and sample below, as an Appendix.
One final preliminary comment is in order. A familiar fact of depressive behaviour is a significant slowing of speech patterns, in some cases extending to mutism and non-movement. This has long been recognised in the descriptive psychopathology of severe depression, where it has been described as 'psychomotor retardation'. Some psychiatrists take psychomotor retardation to be the critical feature of 'melancholic depression' (Taylor and Fink 2006). Non-intuitively, it can exist alongside, or alternate with, another behaviour pattern: severe anxiety and agitation without an object. One or both patterns were present in all of the severe depression cases in this study. In reading the excerpts below, it is helpful to have a sense for the distinctive slowed pace of these conversational exchanges. Prompts, repetitions and patience were often necessary to facilitate communication; significant pauses were common. The extracts quoted below are best read with this held in mind.

Agility and inability
Ultimately, the thesis that we wish to advance concerns a distinctive sort of temporal incapacity that is associated with decision-making among severely depressed patients. But before we try to hone in on just what this inability amounts to, it is worth reminding ourselves of the extraordinarily rich and broad set of temporal agilities that manifest themselves in ordinary human experience. Consider the course of a routine workday. We are aware of the passage of time; we 'manage' our time more-or-less effectively, allocating available time to different projects and purposes. We 'read' the clock and the calendar and the diary; we temporally coordinate with others to arrange meetings and avoid collisions. We use time, save time, waste time, and so on, thereby treating time as a kind of resource. We also engage in all manner of larger-scale temporal activities: We plan for the future (whether over the next few moments or the next few days or over much longer periods); we may recount or reflect on our own individual past or a broader collective historical past. We worry; we hope; we anticipate. We recognise the fact of change over time.
The foregoing list is anything but exhaustive; we include it here in the first instance simply as an indicator of both the pervasiveness and the diversity of our temporal experience. But these reminders also help to exhibit a second important point. The temporal structures enumerated here in each case involve distinctive temporal agilities characteristic of the ordinary life of healthy persons. In what follows, we shall be concerned not with time as an abstract physical or metaphysical structure, but with time as a structure of experience that itself implicates temporal agilities in this sense.
Phenomenological descriptions of depression have often involved claims about some kind of collapse of the temporal dimensions of experience. Straus (1928) claimed that depression "brings internal time to a standstill;" von Gebsattel (1928) held that in melancholia, "the future is shut off;" Binswanger (1960) argued that depression involves a "disturbance of protention". More recently Fuchs has argued that melancholic depression is "triggered by a desynchronisation from the social environment" and results in temporal distortions that he variously describes as "reification of time," "stagnation of personal time," and as a "freezing of selftemporalisation and a loss of contemporality." (Fuchs 2001(Fuchs , 2005a(Fuchs , b, 2010. Against the background of these sorts of claims (to which we shall have occasion to return), a first striking feature of the data from our study is the surprising degree of temporal agility among even severely depressed patients. That feels like a long time? D4: Yeah. I want to get back to my room and rest and then, after that … I've got to have a shower.
As a baseline observation, then, we must be clear that the depressed patients in this study were in a variety of ways temporally agile in the sense of being oriented in time, navigating time and coordinating temporally with others. Despite all this, there were unmistakeable traces of some kind of disorder or impairment of temporal structures in experience, particularly as regards the future.

F. Interviewer:
Can you see options? Can you see options for your future? D1: No.

G. Interviewer:
What's your future? What can you see?

D2:
Not a lot.
In light of these preliminary observations, we can frame the task of second-person hermeneutic phenomenology as follows: What interpretation of the data from these interviews best articulates the complex combination of temporal agility and temporal inability characteristic of the experience of these research subjects?
As a first approximation one might answer this question in terms of hopelessness, a condition that has often been associated with depressed patients. 2 Hope, of course, is itself a temporally inflected psychological condition; it is a way in which we project ourselves onto a better future. In this sense, collapse of hope can be understood as a temporal inability.
Taken on its own, however, this way of identifying the phenomenon proves to be insufficiently fine-grained and fails to discriminate the severe from the mild cases of depression. In fact, a significant number of the severely depressed patients in the study did give voice to expressions of hope, including the hope that they might recover from their present condition.

H. Interviewer:
How do you think things are going to be in the future? D6: I do not know. I cannot tell you. I do not know. I hope I will recover. I hope one day that I recover, you know.

I. Interviewer:
Do you see a future for yourself? D1: I hope so.

J. Interviewer:
What do you hope for in the months or years to come? D3: Very hopeful that things will come out ok. K. Interviewer: What are you hoping for? D6: That I will get well.

L. Interviewer:
Could you be better? D1: I'd hope to be.

Interviewer:
Yeah. What do you mean? What would be better? D2: Well to be better.
Of course, remarks like these might be interpreted in a variety of ways. In order to explore the phenomenology, it is therefore crucial to develop strategies that would probe beyond mere generic expressions of hope. One technique we employed for this purpose was to ask research subjects to specify their hopes, for example, by imagining how their life might be different in 6 months or a year's time. Here an important differential pattern began to emerge. When prompted to flesh out their hopes for the future, to specify their hope with determinate content, severely depressed patients (in stark contrast to their mildly depressed counterparts) were unable to provide any details. Requests for specificity of hope characteristically met with long measurable silences, or with expressions of inability.

M. Interviewer:
Can you see yourself in a future? D1: Mmmm? Interviewer: Can you see yourself in one year or so? Oh I don't know. I have to get through tonight.

Interviewer:
Does it feel like an eternity? D2: Mmm.

O. Interviewer: [subject repeatedly anticipates being dead by end of day].
You feel your life rushing away, pouring away from you.

Interviewer:
How much time do you think you have left on this earth? D3: Oh, in terms of …no time really.

Interviewer:
And what do you want to happen before your time is up?
[38 seconds without reply.] What do you want to happen, before your time is up? D3: Sighs. I'd like to be well again.
[…] Interviewer: How can you bridge that? How can you build a bridge between how you're feeling-that you're going to be dead by the end of the day-and a possible future? D3: Ah! Interviewer: The possible future you're talking about. How can you bridge it? D3 : Ah, ah, ah! I don't know-that's hard.

P. Interviewer:
What about tomorrow? D5: I don't know. Same as today-nothing.
With this, we are in a position to provide an initial description of an asymmetry that emerged in the study: Among severely depressed patients, and in contrast to their mildly/moderately depressed counterparts, the capacity to specify hopes was significantly compromised.
The next step must be to refine this formulation, using the methods of the study, in order to try to articulate more precisely in what this inability consists. Here, it is important to note that this initial evidence of inability arose in response to an ideation challenge. Research subjects were asked to engage in an act of imagining the future and showed evidence of an inability to do so. But one important lesson from the phenomenological tradition is that explicit ideation is only one aspect or dimension of experience, including temporal experience 3 ; it is therefore crucial to look further in order to identify the larger pattern of which this failure of ideation forms a part.

Temporal abilities and the phenomenology of deliberative experience
Think of a circumstance in which you have had to engage in explicit deliberation about what to do. Perhaps you have to decide whether to agree to the terms of a contract, or whether to make a move, or simply about where to spend your weekend. How, in such circumstances, do you find yourself situated or oriented in time? We distinguished four temporal components of these sorts of ordinary deliberative experiences: 1. In the context of deliberation, I encounter the future as 'open,' as 'yet-to-bedetermined' or 'yet-to-be-realised. ' The events of last weekend are experienced as fixed and unrevisable, but the events of next weekend as yet await determination. 4 2. I experience my present task of deliberation and choice as shaping the future: What I decide now has the potential to determine which among those possible futures is realised. Of course, I also might experience certain constraints (say the lack of funds, or the wishes of my family about how to spend the weekend), but I experience myself as shaping the future by responding in one of (what I experience to be) several possible ways to these constraints. 3. Among the various different possible futures that may (or may not) be realised, I recognise significant normative differences: Some are better/worse than others, along various dimensions of normative difference. Staying home would be cheaper; going to the mountains would be beautiful and good exercise, but not so much fun if it rains. 4. When it is deliberatively necessary to do so I can imagine different future scenarios in some degree of detail. If we are late getting back on Sunday evening, will I have time and energy to be sure that I am ready to teach on Monday morning? It is worth emphasising that the exercise of these abilities does not always involve explicit ideation. The first two abilities express something fundamental about the temporal structure of the world in which I find myself faced with a task of choice. But they do not typically involve explicit conscious representational content. The fourth, by contrast, does involve specifically ideational skills: I imagine different possible outcomes and use this imaginative exercise as one strategy for reaching a decision about which outcomes to pursue and which to avoid. The third ability occupies an intermediate position: In explicit deliberation about what to do, recognition of normative differences among possible futures often involves explicit conscious representation of, for example, the difference between a dry and a wet day in the mountains. But we can also recognise and respond to such normative differences without explicitly representing them, as Wrathall (2013) has recently argued.
In our iterative tagging of the interview transcripts, we used these four temporal abilities to sort extracts from the transcripts relating to treatment decisions that were live for the subjects. We were looking for expressions reflecting these various temporal abilities associated with deliberation. A striking pattern emerged. In the interviews with mildly/moderately depressed research subjects, the interview protocol prompted responses that showed traces of all of these components of ordinary deliberative temporality. But among the severely depressed patients, these traces were largely absent when it came to decisions about treatment. In documenting this result, we begin with a few examples from the mild/moderate sample: Q. N2: I don't know whether-not just CBT-I don't know whether anything can make me change the way I see it.

Interviewer:
That's what I suppose I'm asking, whether or not you feel that anything can change it? N2: But obviously I'm prepared to give it a try.

Interviewer:
Why? … N2: I just don't want to be so miserable all the time, you know … Interviewer: But if it's the sort of thing that can't change … then why would you be interested in trying it [CBT]? N2: Well I don't know if it isn't yet, do I, until I try? So … Interviewer: You're going to have to try? N2: I have to try and find out, you know. I mean I'm prepared to give it a go, that's the thing. I mean what else can you do? I don't want to just sit here wallowing all the time, you know, just getting down and finding I can't get a way out of it. It's suffocating.

Interviewer:
So you would like things to change? N2: Oh definitely, yeah, definitely.

Interviewer:
But what you would like to change is that feeling … N2: Feeling all the time, yeah, doom and gloom.

Interviewer:
Doom and gloom? You'd like to change that? You also have this very strong doubt whether that's the sort of thing that can be changed? N2: Yeah, yeah, I mean it's a mindset. It's …the way I've been thinking for so long, you know you just get into patterns, don't you? And it would be interesting to see whether the patterns can be changed. Here, abilities (1), (2), (3) and (4)  Well I was given antidepressants but I'm not really comfortable taking antidepressants because I used to have a serious drug addiction to cannabis, and to me just in general being on drugs doesn't really help anything, and even if it's antidepressants I've heard a lot of stories of people who end up becoming addicted to antidepressants, and I don't want to be the type of person that's reliant on a drug. Here, abilities (1) and (3) are evident.
In these extracts, along with many more like them, the mildly/moderately depressed research subjects systematically show the temporal phenomenology we detailed above. They find themselves situated in an unfolding future that is not (yet) fixed or determined, in which significant normative differences can be identified, and where those differences are themselves linked back to the current decision.
In the severe group, by contrast, it is the absence of these temporal structures that predominate when it came to treatment decisions. By way of example, consider the way D4 struggles to imagine the future in any degree (U) and does not report finding any normative variance among future options (V), while D5's comportment suggests neither openness to the future nor that D5's decisions could shape it (W):

U. Interviewer:
So at the moment what do you think is the right thing for you? D4: I don't know really. They wouldn't let me go home, would they, unless I had a load of support? That wouldn't make any difference.

Interviewer:
So it doesn't make a difference? D4: No. V. Interviewer: So, you're in hospital at the moment and there's the decision about whether to stay here or not stay here. There's this decision about whether to stay in hospital.

Interviewer:
How do you see yourself in time to come, like in the next few weeks or months? What do you hope for? D4: Don't know.

Interviewer:
You don't know? D4: I don't know whether I've got a few weeks.

W. Interviewer:
What would happen if you stopped the insulin? D5: I don't know. All these questions I can't answer.

Interviewer:
You know about diabetes. You've been diabetic for a long time. You used to do your own insulin when you were better.

D5:
When I was well, that was. Not now.

Interviewer:
Because now? D5: I've come to a standstill.

A methodological challenge
Before going further, we need to address a methodological challenge that is endemic to our research strategy. To bring the issue into focus, consider a pattern that recurred several times in two separate interviews with Research Subject D4. Here are some examples: X. Interviewer: You don't hope for anything? D4: No not really.

Interviewer:
Why is that? D4: I don't know. It's just that I don't think I've got a future.

Interviewer:
Can you think of a future at all? D4: Not really, no.

Y. D4:
Well I have no future have I? I can't see a way ahead. Z. D4: I can't see any future for myself, to be honest.
Statements of this kind were common in the interviews with severely depressed research subjects and were very much in keeping with the larger pattern we have been trying to articulate. Like the other severely depressed patients, D4 seemed to lack the sense that his present treatment choices might make a normative difference to his yetto-be realised future. But how exactly should we understand D4's claims here? In particular, how complete is the collapse of D4's hope and sense for a normatively discriminable future? Taken at face value, D4's reports might be understood to suggest that hope and temporality have collapsed entirely.
But to draw such a conclusion would be a mistake, both methodologically and substantively. D4 is struggling to articulate the character of his own experience when faced with a serious decision. This sort of phenomenological reporting is challenging at the best of times (How do we find the right vocabulary? how do we ensure that we are not confabulating, etc.). And D4's circumstances are very far from optimal. He is suffering from a debilitating mental disorder and is trying to report on what must be a very painful matter indeed. It would be naïve to assume that a literal interpretation of his reports accurately captures the character of his lived experience. These sorts of considerations have sometimes been used to motivate a general skepticism over phenomenological methods. 5 After all, if we can't trust first personal reports, then what other access can we have to the experience of others? But this sort of skepticism is itself naïve. A probing interview by an experienced psychiatric professional is a powerful tool for phenomenological analysis, and there is good reason to believe that we can use that tool to learn about the experience of patients. Nonetheless, it would be a methodological error simply to take the reports of research subjects at face value. 6 Our own strategy for dealing with this methodological challenge was to deploy the methods of critical hermeneutics. Second-person phenomenology is an essentially interpretative undertaking, and like any interpretative exercise, it must be critical with regard to its texts and sources. Accordingly, we drew on the familiar maxims of hermeneutic investigation in developing an interpretation of patient reports: Consider the speech acts of the patient not in isolation but in context, as parts of a larger hermeneutic whole; be on the lookout for patterns of consistency and inconsistency in that larger context; where inconsistency arises, consider how that inconsistency can be rendered meaningful. In sum: Find an overall interpretation that accounts both for the content of the reports and for the inconsistencies and tensions within them. Consideration of this overall content can begin with the explicit content of the reports from the research subject, but it should also take into account other features of the situation. In particular, we must consider not only what the patient says but also what he shows by his comportment. It should go without saying that no application of these hermeneutic techniques can yield anything like a final proof as regards the experience upon which the research subject is reporting, but they do provide a systematic technique for dealing with the endemic methodological challenge; moreover, the results turn out to be revealing.
Consider how this methodology applies in the case of D4. As we have seen, D4 explicitly denies that he can think about the future and that he has any hopes. But consider now the context in which these denials are embedded.

AA D4:
I can't see any future for myself, to be honest. Just more of the same … BB Interviewer: Can you think of a future at all? D4: Not really, no.

Interviewer:
What image do you have in your mind when you think about the future? D4: Nothing really, just the same struggle to keep going. CC Interviewer: Yeah, but you said "I don't think I can see a future, given that I know what I know now".

Interviewer:
You said that didn't you? D4: Yeah.

Interviewer:
Tell me what you mean by that because I'm not sure I understand it and I'm trying to just understand it. D4: Well Is it just like a feeling that you don't know? D4: Yeah. Can I go back to my room? Interviewer: Yeah sure, but look, we haven't quite finished, ok? We'll be finishing very soon, but I just want to round up and then we'll get you back to your room, ok, very shortly? Is that alright? D4: Alright.
Once we begin to take account of the broader context, D4's reports take on a rather different texture. He says that he can't see any future for himself, but he immediately goes on to report on the future that he sees: 'just more of the same', 'just the same struggle to keep going'. That future itself is clearly one that involves normatively significant change: Bringing an end to the interview is preferred over continuing it (and who would blame him, the reader might ask!). Even while he reports that he can't find a way ahead, he nonetheless projects himself into a future that involves getting back to his room. Moreover, we can clearly see that he is here temporally coordinating ('synchronising') with the interviewer-effectively negotiating an agreement about how much longer the interview will continue. Any adequate interpretation of D4's reports of temporal collapse must somehow account for the temporal abilities that he continues to exhibit in his comportment.
What is the import of all this? Methodologically, it serves to illustrate the critical dimension of second-person hermeneutic phenomenology. Reports from research subjects are to be taken as data, and our epistemic stance towards that data must be carefully calibrated. Our working stance is that phenomenological reports do indeed tell us something about a patient's deliberative experience. Moreover, as with any experimental data, our confidence increases to the extent that we find the same patterns replicated across discrete cases. At the same time, we recognise that such data does not provide a perfect window upon the experience of the research subjects. The D4 case serves as a salient reminder of that fact, but it also suggests a strategy for coping with the methodological challenge that ensues. For every tension or contradiction in the reports from a research subject can in turn become a critical standard for the interpretation of the data: The best overall interpretation of the data must be one that can make sense of what otherwise appears to be an inconsistency.
The tensions within D4's utterances and between his utterances and comportment are not only methodologically important; they also have substantive significance, particularly for assessing and refining some of the existing phenomenological models of depression. To take just one example, consider Thomas Fuchs' recent account of the impact of severe melancholic depression on the usual temporal structures of experience. Fuchs has argued that under conditions of severe melancholia, the patient suffers "a complete desynchronization between individual and environment" (Fuchs 2001: 9), including what Fuchs describes as "a complete desynchronization or uncoupling from intersubjective time" (Fuchs 2005b: 196). Our data raise doubts about the accuracy of Fuchs' model. The profoundly depressed patients in our study must certainly be classified as having severe melancholia (see Section 1); nonetheless, they exhibited a significant degree of temporal coordination with objects and persons in their environment. As we have seen, D4 coordinates temporally with the interviewer, even while denying that he can project himself into the future. We found something similar in other cases. The most severely depressed research subject in the study (D6) was a woman awaiting ECT for her depression. D6 suffered from an unusual form of solipsism: Again and again she reported (sometimes quite insistently) that there was no one else in the hospital, perhaps even no one else in the world. But the same patient was perfectly able to reach out and accept a mug of hot tea that was handed to her, a skill that clearly requires temporal coordination with others. We have already seen above that patients have at least some future goals and plans, including those involving coordination with others (arranging another research interview, meeting their spouse). 7 So how best can the tensions in the D4 interviews be explained? The first step is to understand two of D4's reports as an integral pair. As we have seen, he denies that he sees a future, and he reports on a future that involves "more of the same". The key is to see the tension between these two reports as itself integral to the phenomenon that D4 is trying to articulate. Recall the way in which ordinary deliberative phenomenology involves the recognition of normatively significant differences among different yet-to-be-realised futures. When I engage in deliberation, it is not enough to see that time will continue (inside, the clocks continue to tick; outside, the buses continue to run …). It must also be the case that I can identify significant differences among the possible ways that the future will unfold. It is in selecting among those possible futures that the act of deliberation, so to speak, gains traction. When D4 reports that he can see no future for himself, the best interpretation of this claim is that his inability pertains specifically to this dimension of normative variation in his future. 'I can't see a future' means something like, 'The future is normatively flat for me; I can't see a range of possible futures of which some are meaningfully better than others.' Interpreting his report along these lines captures both sides of the noted tension: D4 can see a future, but it is not a meaningful future, not a future that makes any difference, not a normatively differentiated future.
But this by itself is not enough to yield a coherent overall interpretation of the experience upon which D4 is reporting. For we have also seen evidence that D4 does recognise normative differences in his future: He clearly prefers the future in which the present interview ends sooner rather than later and in which he is able to get back to his room. Here, we can best make sense of D4's reports by recognising an important dimension of decision specificity in the temporal inabilities associated with severe depression. It is crucial to recall that D4's expressions of temporal inability were offered in response to a line of questioning specifically pertaining to his pending treatment decision. It was in the context of deliberating about that decision and then reflecting on that deliberation, that D4 expresses the inability to see a future, or to specify his hopes for the future. But when it came to other decisions (e.g. the decision about whether to continue the interview), D4 was able to recognise himself in the future, and indeed to draw normative discriminations among various yet-tobe-realised futures. Here, it is worth repeating a passage from the D4 transcripts that we cited at the outset as an example of temporal agility among severely depressed patients: "Yeah. I want to get back to my room and rest and then, after that … I've got to have a shower." The specific decision at hand here is whether to continue the conversation or whether to have a shower. This might be a trivial decision, or it might be something that matters quite a lot to D4. Either way, we can see that D4 is here exercising the four temporal abilities that we have enumerated: D4 makes normative differentiations about a future that is experienced as open and as something that is shaped by D4's decisions and that can be imagined in some detail.
In legal and ethical discussions of incapacity, decision-specificity has in recent years become an important concept, and a powerful tool for legal reform. There was a time not long ago when a finding of mental incapacity was treated as an overall legal and clinical status, with far-reaching results for the patient involved. A patient who was found to be lacking in mental capacity might be made a ward of the state, for instance, with a consequent loss of any rights to make decisions for him/herself. At the time of writing, this status-based legal arrangement is still in effect in many jurisdictions. In reaction against the status-based approach, activists and jurists have pressed for a functional test which is specific to the decisions with which the patient is faced. Where such pressure has been successful-as it was in case of the Mental Capacity Act 2005 in England and Wales-this has created a legal space in which patients might be found incapable of making some decisions but retain the right to make others. The movement towards decision specificity in the legal conceptualisation of incapacity has in large part been driven by legal and ethical considerations, but our data indicate that decision-specificity maps onto an underlying clinical reality. Even in states of severe depression, where we see global transformations of experience, decisionspecificity can be shown. In the case of D4, the best overall interpretation is that his temporal inabilities exhibit decision-specificity. To further test this hypothesis, we asked four experienced psychiatrists at the Institute of Psychiatry, London, to review the full 18-page transcript of our interview with D4. All agreed that D4 exhibited capacity for the decision to participate in the research study, but not for the treatment decision. What is the nature of a research decision compared with a treatment decision? Conventionally, the distinction is taken to be one of personal care (treatment) versus impersonal knowledge (research) (Appelbaum et al. 1987). We think that further work looking at the self-implicating features of treatment decision-making versus the other-implicating features of research and how these map to decision-making capacity in depression would be valuable. Other variables-short-term and long-term aspects of the two decisions, or seriousness and triviality of them-would also have to be explored but our data did not provide conclusive evidence on these matters.

Legal and clinical relevance
Both in legal and in research contexts, the concept of decision-making capacity has characteristically been elaborated in terms of a four-abilities model. The exact enumeration of these abilities has been subject to some variation. The widely used MacArthur Competence Assessment Tool-Treatment (MacCAT-T), for example, assesses a patient's ability: What is common to both models is both a theoretical and a practical strategy. Theoretically, decision-making capacity is analysed as a decision-specific threshold that itself comprises a set of four constituent abilities. In other words: A particular patient at a particular time either has or lacks the capacity to make a particular decision. The presence or absence of that capacity is in turn understood in terms of these four abilities, which are treated as individually necessary and collectively sufficient for capacity. In practice, this approach means that the assessment of capacity (and the articulation of a clinical or legal judgement of capacity) can be broken down into stages, with individual assessment of the component abilities feeding in to an overall judgement of competence. This approach lies at the heart of the MacArthur design and is explicitly provided as a model for capacity assessment in codes of clinical practice.
There can be no doubt that the four-abilities models have proven value in guiding and ordering assessments of capacity. Tests such as the MacCAT-T are now widely validated, and legal standards such as that in the MCA have been tested in the courts. But the assessment of capacity remains, by universal agreement, as much an art as a science. Tools like the MacCAT-T provide a framework, but ultimately the assessment comes down to the art of interpreting the mental state and condition of a patient. In so doing, some of the constituent abilities in the four-abilities model prove to be much more straightforward to assess than others. Some abilities can be assessed simply by soliciting overt performances from patients. But notoriously, assessment of other abilities requires considerably more interpretation. In the context of both law and clinical practice, it is the ability to use or weigh (or to "appreciate" and "reason," in the terminology of the MacArthur approach) that has proven to be the most difficult to interpret and assess. What does it actually mean to use or weigh information in reaching a decision? And how is this ability to be assessed in circumstances of mental disorder? These issues have been particularly challenging in the context of depression. Legal judgements on decisionmaking capacity involving depression are sparse compared with other psychiatric disorders, and the MacCAT-T articulation of appreciation has proven difficult to apply with severely depressed patients. 10 It is precisely here that the results of the present study can be put to work-both in further articulating the concept of decision-making capacity and in the practical enterprise of assessing capacity in the context of depression. Conceptually, we propose that a range of distinctively temporal abilities are themselves implicit in the capacity to make decisions for oneself. The idea of deliberative 'using or weighing' has often been represented with the image of the scales, but the experience of deliberation must be seen to involve a complex ability to situate oneself in an extended and asymmetrical temporal landscape. The ability to 'use or weigh' can usefully be understood as involving a distinctive way in which we orient ourselves in time: drawing on our past in order to project ourselves into the future by way of decisions taken now. In order for that complex activity to be possible for me (i.e. in order for me to have decision-making capacity), I must have abilities that are specifically temporal.
To see this, the key point to recognise is that the temporal abilities we distinguished in Section 3 (above) are more than simply temporal features of ordinary deliberative experience; they provide the temporal context in which deliberation and decision-making is intelligible at all. Decision-making is an activity that is concerned with shaping the future; if I have lost the ability to experience the future as "open" or "yet-to-be-determined," then decision-making can no longer make any sense for me. If the available futures all manifest themselves as normatively 'flat,' then there can be no sense in asking me to choose one over another. And if I have lost the ability to see my present actions as potentially determining which future is realised, then I no longer genuinely understand what it is for me to make a choice. In this sense, temporal abilities are conditions on the possibility of decision-making capacity. 11 Our study has shown that these temporal abilities are themselves a significant variable in the population of patients diagnosed with depression. In particular, patients with mild or moderate depression show no evidence of loss of these abilities whereas patients with severe depression do. This result in turn has practical consequences. An effective clinical assessment of decision-making capacity must be guided by an understanding of the constituent abilities that themselves comprise capacity and by the risk-factors associated with particular mental disorders. 12 In probing decision-making capacity, assessors should be alive to the specific risk factors associated with depression, and devise avenues of questioning which can elicit the temporal abilities of the patient. Among populations of severely depressed patients, assessors of the ability to "use or weigh" should not be satisfied with generic expressions of hope, or even the loss of it, but should probe the patient's ability to project himself into yet-to-be-realised futures which themselves exhibit normatively significant differences, while also recognising that that very ability can be decisionspecific. In future work, we plan to develop these assessment strategies. 11 Our claim in this context is that the first three temporal abilities distinguished above are essential for decision-making capacity. It is an open question whether the fourth, specifically ideational skill is necessary. Decision-makers are typically capable of envisioning different possible futures in some detail, and for certain kinds of decision, this ability may well be essential. But there are other contexts and strategies of decision-making that may not lean so heavily on this particular ability. We may, for example, decide for one option over another on the basis of a moral conviction, or a feeling of unease about one of the two options. This may be sufficient to underwrite decision-making capacity in that particular case, even if we find ourselves unable to imagine the consequences in any degree of detail. 12 We have on occasion heard the argument put forward that assessments of mental capacity ought to be "blind to diagnosis" in the interest of nondiscrimination. This claim raises a number of thorny issues that go beyond the parameters of the present discussion. Suffice for present purposes to say that we do not share this view, nor do we believe that it accords with the requirements of the MCA, which permits a finding of incapacity only where the lack of decision-making ability is itself caused by an impairment of or disturbance in the mind or brain (MCA §2.1.). We believe that this provision of law requires assessors to be sensitive to diagnosis, if for no other reason than for considering the issue of causality. But we would also argue that responsible capacity assessment always involves a process of getting to know the patient and that familiarity with diagnosis forms an essential part of this process. the patients were detained under the The Mental Capacity Act (2005)-the law in England and Wales authorising compulsory treatment for mental disorder-a situation that is common for people with severe depression who, though very ill, often do not refuse treatment. Three out of seven were Black African or South Asian, reflecting the ethnic diversity of South London. Three out of the seven were women. The average age was 66 years (range, 36-77 years). All were fluent English speakers with education histories and attainment that, though varied, implied intelligence not significantly less than average. None of the patients were able to work when interviewed, but each one had previously done so. Four of the seven patients were interviewed in their homes, and three were interviewed in psychiatric hospital wards. Each one faced significant decisions, at the time of interview, about treatment that ranged from medication (psychiatric or medical), electro-convulsive therapy (ECT) to whether to be receiving home or hospital care.
Characteristics of the mild depression group Six patients (N1-6) participated who had been referred by their family doctors (GPs) to a psychological treatment centre with concerns expressed about low mood. When interviewed by Gareth Owen all met ICD-10 criteria for mild or moderate depressive episode. Self report scores on the Beck Depression Inventory (BDI-II) ranged from 17 to 36, indicating that the patients were significantly symptomatic. Three of the six were women. Average age was 39 (range 22 to 61 years). All were fluent in English, and all but one was in work. Each patient had been assessed by a clinical psychologist as suitable for psychological treatment for depression where the predominant model in use was cognitive behavioural therapy ("CBT"). The patients were waiting to begin treatment and, when interviewed, faced a decision about proceeding.